Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2508.14345 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914000296476672 |
|---|---|
| author | Rios, Gaston Gustavo Bianco, Pedro Dal Ronchetti, Franco Quiroga, Facundo Stanchi, Oscar Ahón, Santiago Ponte Hasperué, Waldo |
| author_facet | Rios, Gaston Gustavo Bianco, Pedro Dal Ronchetti, Franco Quiroga, Facundo Stanchi, Oscar Ahón, Santiago Ponte Hasperué, Waldo |
| contents | Sign Language Recognition (SLR) models face significant performance limitations due to insufficient training data availability. In this article, we address the challenge of limited data in SLR by introducing a novel and lightweight sign generation model based on CMLPe. This model, coupled with a synthetic data pretraining approach, consistently improves recognition accuracy, establishing new state-of-the-art results for the LSFB and DiSPLaY datasets using our Mamba-SL and Transformer-SL classifiers. Our findings reveal that synthetic data pretraining outperforms traditional augmentation methods in some cases and yields complementary benefits when implemented alongside them. Our approach democratizes sign generation and synthetic data pretraining for SLR by providing computationally efficient methods that achieve significant performance improvements across diverse datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_14345 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | HandCraft: Dynamic Sign Generation for Synthetic Data Augmentation Rios, Gaston Gustavo Bianco, Pedro Dal Ronchetti, Franco Quiroga, Facundo Stanchi, Oscar Ahón, Santiago Ponte Hasperué, Waldo Computer Vision and Pattern Recognition Machine Learning Sign Language Recognition (SLR) models face significant performance limitations due to insufficient training data availability. In this article, we address the challenge of limited data in SLR by introducing a novel and lightweight sign generation model based on CMLPe. This model, coupled with a synthetic data pretraining approach, consistently improves recognition accuracy, establishing new state-of-the-art results for the LSFB and DiSPLaY datasets using our Mamba-SL and Transformer-SL classifiers. Our findings reveal that synthetic data pretraining outperforms traditional augmentation methods in some cases and yields complementary benefits when implemented alongside them. Our approach democratizes sign generation and synthetic data pretraining for SLR by providing computationally efficient methods that achieve significant performance improvements across diverse datasets. |
| title | HandCraft: Dynamic Sign Generation for Synthetic Data Augmentation |
| topic | Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2508.14345 |